We propose to develop a second generation mathematical model and computer systems program dealing with the impact of screening on the natural history of and mortality from breast and cervical cancer. The program objectives are: 1) To extend an existing first generation model, 2) to develop statistical estimation techniques in order to furnish the inputs necessary for the exercise of the model described above; 3) to design screening programs so that data suitable for the goals of 2) are obtainable; and 4) to apply these methods to existing bodies of data such as the British Columbia Cervical Cancer Screening Projects and the Breast Cancer Demonstration Project in order to obtain insights into the natural history of cervical and breast cancer. In turn, these insights will allow us to design screening programs which utilize limited resources efficiently.